With LLM-based tools that inherently rely so much on the semantics of language, I wonder if there will be differences in code generated for the "wanted board" in the "Wasteland", compared to the "task list" in the "public square" or the "wish list" in the "Utopia".
Infinite Jest describes a very similar (fictional) development, albeit with network TV. As viewers leave, content producers are ever-more desperate to monetize remaining traffic, which worsens the experience and drives more viewers away, creating even more desperation to monetize... a vicious cycle.
@dang would there be any possibility of creating a view that hides posts and comments by accounts newer than, say, Jan 1 2026? Similar to how https://news.ycombinator.com/classic works (only showing votes from the oldest accounts)?
I know this is unfair to prospective new community members, but I'm unsure of other good methods to filter out AI bots at scale. Would certainly welcome other ideas.
For what it's worth, your original comment is a little hard to parse, particularly because you say "our pilot deployment" which makes it sound like you were involved in deploying the cameras. Combined with your realpolitik comment about knowing your neighbors want them, I think several people are confused about your opinions and what you ended up doing to fight the cameras.
I wonder how much our understanding of past language is affected by survivorship bias? Most text would have been written by a highly-educated elite, and most of what survives is what we have valued and prized over the centuries.
For instance, this line in the 1800s passage:
> Hunger, that great leveller, makes philosophers of us all, and renders even the meanest dish agreeable.
This definitely sounds like the 1800s to me, but part of that is the romance of the idea expressed. I wonder what Twitter would have been like back then, for instance, especially if the illiterate had speech-to-text.
There's also a lot of historical writing out there that's more or less the shorthand scribblings of shopkeepers, foremen, and low-level clerks, so it's not all flowery prose. There's even surviving Egyptian hieroglyphics that are more or less just work logs, and they're quite different than the painted ones in the tombs. Then there's the graffiti that's all over Pompeii.
So about 1 in 10? Doesn’t seem that terrible to me. Especially when many of them are in response to questions about his work, and he’s answering with a link to a different post.
Fair enough - I didn’t turn the sound on for the video. The premise of the sketch seemed to fit the rest of the slimy stuff in the feed, which is why I figured that video in particular got recommended.
I mostly included it because of the absurd question Meta suggested I ask their AI.
Interesting. I wonder if the phrasing as "do you ever use this platform" leads to this result. I could definitely believe more 30-49 year olds log in every so often for Marketplace etc., but would expect DAU to be lower than 50+. But maybe that's just more of the same bias you describe.
Well, I haven't really used any Meta platforms for at least 5 years, so I don't think that's how they're deciding what to serve me.
I could definitely believe that I used to click on more pictures of girls than boys back in high school and college when I actually used Facebook. But they would have been real pictures of people I was friends with.
To your point, I'm sure if I used the product more, the algorithm would get "better" according to what I engaged with.
reply